China Economic Review 11 (2000) 171 ± 188
The China money puzzle: will devaluation of the yuan
help or hurt the Hong Kong dollar?
Shang-Jin WEIa, Ligang LIUb, Zhi WANGc,*, Wing T. WOOd
a
Harvard University, 79 John F. Kennedy Street, Cambridge, MA 02138, USA
b
World Bank, 1818 H Street NW, Washington, DC 20433, USA
c
Economic Research Services, United States Department of Agriculture, Room 5141,
1800 M Street, NW, Washington, DC 20036-5831, USA
d
University of California-Davis, One Shield Avenue, Davis, CA 95616, USA
Abstract
A Chinese yuan devaluation could affect the stability of the Hong Kong dollar. This paper studies two
linkages. The first is that trade balance effect is studied through a computable general equilibrium (CGE)
model. The result shows that the net change in Hong Kong's foreign reserve after a yuan devaluation is,
in fact, negligible. The second is that psychological effect is studied by a survey of financial market
participants. In spite of the small trade balance effect, all respondents believe that a yuan devaluation
would lead to a panic selling of Hong Kong assets. Therefore, a yuan devaluation is bad for the Hong
Kong dollar primarily through market psychology. D 2000 Elsevier Science Inc. All rights reserved.
JEL classification: F31; F47; O53
Keywords: Chinese yuan and Hong Kong dollars; Trade balance; Psychological panic; Computer general
equilibrium model
1. Introduction
Against the background of domino devaluations of most Asian currencies around China,
the relative stability of the Chinese yuan, and the repeated pledges made by the Chinese
leaders not to devalue the yuan, have attracted attention. On several occasions, US Treasury
officials and International Monetary Fund officials have openly praised the Chinese decision
to hold the yuan steady.
* Corresponding author. Tel.: +1-202-694-5242; fax: +1-202-694-5793.
E-mail address: [email protected] (Z. WANG).
1043-951X/00/$ ± see front matter D 2000 Elsevier Science Inc. All rights reserved.
PII: S 1 0 4 3 - 9 5 1 X ( 0 0 ) 0 0 0 1 4 - 6
172
S.-J. Wei et al. / China Economic Review 11 (2000) 171±188
Is this a rational policy? Most will say that it is a good policy for the Asian region and the
world community. But is it a sensible policy for China's self-interest, leaving aside possible
soft ``credits'' it may get from foreign governments and international organizations? The
answer to this question will help us to understand how committed the Chinese government is
to the nondevaluation, and how long they will keep the commitment.
Recent computable general equilibrium (CGE) studies (Noland et al., 1998) reveal that the
Asian devaluations have a clear negative effect on the Chinese economy through two
channels: China's exports to these economies will decline, and more importantly, China's
exports to third-country markets such as the US and Japan will suffer because the Chinese
goods are becoming less competitive.1 The study further demonstrates that a 3±6% real
devaluation in the Chinese currency (a relatively modest change) can restore the competitiveness of China's external trade position ex ante.
Should China undertake the devaluation? The answer depends not just on how it will affect
China's direct exports and imports, but in an important way, on how it will affect the Hong
Kong dollar. For many politically important reasons, Hong Kong's prosperity is of vital
concern to the leaders in Beijing since the reversion of the former British colony back to
Chinese rule on July 1, 1997.2 For many in the international financial community, the stability
of the Hong Kong dollar has become a symbol of the health of Hong Kong's economy.
Hence, how the Hong Kong dollar will be affected is a crucial consideration when the
Chinese leaders' contemplate a possible devaluation of the RMB.
What is the effect of an RMB devaluation on the Hong Kong dollar? Some observers (e.g.,
Hu, 1998) have said that an RMB devaluation would cause a severe run on the Hong Kong
dollar, possibly making the peg indefensible. Other observers have claimed that an RMB
devaluation would trigger another round of devaluations of other Asian currencies that would
reduce the competitiveness of Hong Kong exports, and ultimately reduce its reserve as well.
We will call this negative view the psychological factor. However, the negative outcome is
not a certainty because there is an important offsetting factor based on the fact that Hong
Kong is an important entrepot for trade with China (Wei & Zeckhauser, 1998; Hu et al.,
1998). China's indirect exports and imports through Hong Kong are about 20±30% of
China's total exports and imports during recent years. A recent paper by Feenstra et al.
(1998), found the markup on Chinese exports through Hong Kong to be about 27%, and the
markup on Chinese imports through Hong Kong to be negligible. Since the RMB devaluation
would increase Hong Kong's entrepot exports because of its positive impact on China's
exports, the value-added in the entrepot export would increase Hong Kong's foreign
1
However, China will obtain intermediate inputs and semi-processed products at lower cost from the affected
neighboring countries, such as South Korea, for its exports in processed trade sectors. Processed trade has already
accounted for more than half of China's total exports in 1995. Thus, this effect might mitigate some negative
effects caused by the crisis.
2
The obvious important reasons include, one, the continued prosperity of Hong Kong will be a clear
demonstration to Taiwan of China's commitment to the principle of ``one country, two systems;'' and, two, the
faltering of Hong Kong's economy will reflect badly on China's ability in economic management, a bad signal
to the international community for a country that has come to rely increasingly on foreign direct investment
(Wei, 1988).
S.-J. Wei et al. / China Economic Review 11 (2000) 171±188
173
exchange reserve and enable Hong Kong to defend its currency better. Furthermore, because
the value-added in Hong Kong's entrepot imports for China is negligible, the loss in reserve
due to a decline in China's indirect imports through Hong Kong would be small. Realizing
this connection, currency traders should buy, rather than sell, Hong Kong dollars when the
RMB devalues. In short, devaluation of the RMB creates a positive entrepot effect.
However, there are negative direct trade effects that directly act against the positive
entrepot effect. An RMB devaluation would also reduce Hong Kong's direct exports to China
(and possibly to elsewhere), and increases its imports from China. So the net effect of an
RMB devaluation on Hong Kong's trade balance and, hence, its foreign reserve position is
not clear without careful quantification. We will call this ambiguous net effect the trade factor.
Surprisingly, we are not aware of any quantitative study that addresses this factor.
This paper examines the trade factor and the psychological factor on the Hong Kong dollar
if the RMB were devalued. The trade factor is quantified by using a CGE model to calculate
the effect of an RMB devaluation on Hong Kong's external trade and on Hong Kong's ability
to defend its currency. The psychological factor is assessed by using a small survey of
currency and equity traders in Hong Kong on their reactions to a possible RMB devaluation.
The paper is organized as the follows. Section 2 provides a description of the CGE model
used in this study. Section 3 presents the simulation results and related discussions. Section 4
discusses the small survey of traders. Section 5 concludes.
2. Structure of the CGE model
The model used in this study is a multisectoral, multicountry CGE model and is part of a
family of models that have been used widely to analyze the impact of global trade
liberalization and structural adjustment programs (Wang, 1997a). It focuses on real trade
flows, trade balances, world prices, and real exchange rates. Like most CGE models, it does
not consider financial markets, interest rates, or inflation; and so cannot be used to analyze the
direct impact of the crisis on financial markets.
Detailed algebraic specification of the model can be found in Wang (1997b) and Noland et
al., (1998). Here we provide an intuitive description of its structure. The model includes 18
regions,3 each with 14 sectors4 and five primary factors of production: agricultural land,
natural resources, capital, unskilled labor, and skilled labor.5 The regions are linked by
3
The regions in the model are US, Canada, Mexico, European Union, Oceania, Japan, Korea, Taiwan, China,
Hong Kong, Indonesia, Thailand, Philippines, Singapore, Malaysia, South Asia, Latin America and the Caribbean,
and the rest of the world.
4
The sectors are agricultural products, processed food and beverages, forestry and fisheries, mining, energy,
textile and apparel, light manufacturing, industrial intermediates, motor vehicles and parts, other transportation
equipment, electronics, machinery, housing and construction, and services.
5
Skilled workers are defined as International Labor Office (ILO) International Standard Classification of
Occupations (ISCO) occupation groups 0 ± 2 (professional, technical and related workers; administrative and
managerial workers). The remainder, ISCO 3 ± 5 (clerical and related workers; sales workers; service workers),
ISCO 6 (agricultural workers), and ISCO 7 ± 9 (production and related workers, transport equipment operators, and
laborers) are classified as unskilled.
174
S.-J. Wei et al. / China Economic Review 11 (2000) 171±188
commodity trade. Within each region, the model solves for domestic commodity and factor
prices that equate supply and demand in all goods and factor markets. The model also solves
for world prices equating supply and demand for sectoral exports and imports across the world
economy. In addition, for each region, the model specifies an equilibrium relationship between
the balance of trade (in goods and nonfactor services, or the current account balance) and the
real exchange rate (which measures the average price of traded goods Ð exports and imports
Ð relative to the average price of domestically produced goods sold on the domestic market).
An exogenous change in a particular region's exchange rate will reverberate across the world
economy, affecting the aggregate trade balances and/or real exchange rates of all 18 regions as
they adjust their trade flows and structures of production to achieve a new equilibrium.
In each regional model, there is one competitive firm in each sector, which produces only
one product. The production is characterized by two-level nesting of constant elasticity of
substitution (CES) functions. At the first level, firms are assumed to use two types of inputs: a
composite primary factor and an aggregate intermediate input according to a CES cost
function. At the second level, the split of intermediate demand is assumed to follow Leontief
specification, therefore, there is no substitution among intermediate inputs. Technology in all
sectors exhibits constant return to scale implying constant average and marginal cost.
Products from different regions are imperfect substitutes (the Armington assumption), and
the private household in each region maximizes a Stone±Geary utility function over the 14
composite goods, subject to their budget constraints yielding to the Extended Liner
Expenditure System (ELES) of household demand functions. Household savings are treated
as demand for future consumption goods with zero subsistence quantity (as in Howe, 1975).
An economy-wide consumer price index is specified as the price of savings. It represents the
opportunity cost of giving up current consumption in exchange for future consumption (Wang
& Kinsey, 1994). Government spending and investment decisions in each region are based on
Cobb ±Douglas utility functions, which generate constant expenditure shares for each
composite commodity. In each region, firm intermediate inputs, household consumption,
government spending, and investment demand constitute total demand for the same
Armington composite of domestic products and imported goods from different sources. A
two-level nested CES aggregation function is specified for each composite commodity in
each region. The total demand is first divided between domestic produced and imported
goods, then the expenditure on imports is further divided according to the geographical origin
under the assumption of cost minimization. Complete trade flow matrices for all trade
partners are part of the model solution. To distinguish between Hong Kong's re-exports and
Hong Kong's domestically produced goods for exports, all Hong Kong's re-exports are
allocated back to their original source countries in the model's base year data, therefore, in the
simulation of the model, the solutions of bilateral trade flows between Hong Kong and all its
trade partners, including China, are only imports for its own consumption and exports from
its own domestic production. In the calculations presented in Section 3, we then decompose
China's total trade into its direct trade and indirect trade through Hong Kong.
There is an international shipping industry in the model, and each region allocates a
fraction of the output of its transportation and service sector to satisfy the demand for
shipping which is generated by interregional trade. The global shipping industry has a unitary
elasticity of substitution among supplier sources. This means that the margins associated with
S.-J. Wei et al. / China Economic Review 11 (2000) 171±188
175
this activity are commodity/route specific. In equilibrium, the total value of international
transportation services at the world price equals the sum of the export proportions of the
service sector's output from each region.
The government in each region imposes import tariffs, export subsidies, and indirect taxes,
in ad valorem terms. Tariff and tax (subsidy) rates vary by sector and by destination.
The model determines relative prices within each region and on world markets. Traded and
non-traded goods are assumed to be distinct (and imperfect substitutes) by sector, so changes
in relative world market prices are only partially transmitted to domestic markets. The model,
thus, incorporates a realistic degree of insulation of domestic commodity markets from world
markets, but the links are still important and provide the major mechanism by which the crisis
is transmitted across regions. However, since the model cannot determine inflation, only
relative prices change. The US is specified as the ``reference'' economy, with both its
aggregate price level and exchange rate specified as fixed exogenously. That is, all relative
world prices and trade balances are measured in terms of real US dollars. In addition, the
aggregate consumer price index is fixed exogenously in each region, which defines a ``no
inflation'' benchmark.
The equilibrium exchange rate determined by the model for each region can be
interpreted as the real effective exchange rate (REER) deflated by the ratio of the regional
consumer price index and the US consumer price index. It is important to emphasize that the
exchange rate variable in the model is not a financial exchange rate, since the model has no
assets or asset markets. It represents the equilibrium real exchange rate that is consistent
with a given trade balance.
For each region, the model includes three macro balances: savings±investment, balance of
trade (in goods and non-factor services), and government expenditure±receipts (government
deficit). The three balances are not independent, and the determination of these macro
balances is the subject of traditional macroeconomic models. In terms of our real trade model,
which does not include financial markets or variables typical of macro models, the
determination of these macro aggregates is specified by exogenously determined rules. The
macro adjustment mechanism constitutes the macro ``closure'' of the model.
Since this model is used to explore the impact of changes in real exchange rates on trade
balances, we specify a macro closure that assumes any macro adjustment is spread in a
neutral manner across aggregate consumption, investment, and government expenditure.
Aggregate investment and government expenditure are simply specified as fixed shares of
total absorption in each region (or aggregate regional expenditure, which equals gross
domestic product (GDP) plus imports minus exports). Aggregate domestic savings and
balance of trade (foreign savings) in each region is assumed to adjust endogenously to match
aggregate investment, thus achieving savings±investment equilibrium even with endogenous
changes in the government deficit and the balance of trade.
In the aggregate, as noted above, there is a functional relationship between the balance of
trade (in goods and non-factor services, or the current account balance) in each region and the
real exchange rate. If the real exchange rate depreciates, the price of traded goods increases
relative to the price of domestically produced goods sold on the domestic market. Exports
increase, imports decrease, and the trade balance will improve. Given our assumption that
aggregate investment is determined as a share of aggregate absorption, changes in the trade
176
S.-J. Wei et al. / China Economic Review 11 (2000) 171±188
balance, which directly affect foreign savings, are assumed to have only a partial effect on
aggregate investment in the region. Instead, they lead to an equilibrium adjustment in the
domestic savings rate, which partially offsets the change in foreign savings.
In the base solution, the initial trade balance and exchange rate are assumed to be in
equilibrium for each region Ð that is, the initial trade balance is assumed to be ``sustainable''
and consistent with the initial real exchange rate. In simulation experiments, we change the
exchange rate for a particular region, which changes the equilibrium trade balance both in
aggregate and bilaterally. In the multiregion model, there is a ripple effect, since, as noted in
the previous discussion, a depreciation in one region's real exchange rate implies a relative
appreciation for its trading partners, leading to changes in their equilibrium trade balances as
well. The world model is closed in the sense that the sum of all regional trade balances must
be zero. Thus, we can use the model to see how real exchange rate shocks lead to adjustments
in trade balances worldwide in a consistent framework.
In the simulations, we change the real exchange rate of one region at a time, keeping all
other regional real exchange rates fixed (for a detailed description on the justification of
simulation designs, see Noland et al., 1998). We use the model as a simulation laboratory to
isolate the exchange rate transmission effect. We do not attempt to model a mix of exchange
rate and trade balance adjustments, or to forecast what might actually happen. This real
model, with no money or asset markets, simply cannot be used for such forecasting. The
model does allow us to explore structural adjustment effects such as changes in sectoral
production, exports, and imports. In principle, the model could be linked to a macro model
that includes asset flows and could determine the set of real exchange rates resulting from
some macro shock. Our model could then be used to determine the resulting real trade flows
and regional sectoral structural adjustments.
3. The effect of a yuan devaluation on Hong Kong's trade balance Ð the trade factor
The effects of any shock, RMB devaluation included, on the Hong Kong dollar can be real
(through its effect on Hong Kong's ability to defend its currency, particularly on the level of
its foreign exchange reserve) or psychological (through its effect on market's perception on
Hong Kong authority's commitment to defend the currency even if it were able to do so).
Considering the changes in Hong Kong's foreign exchange reserve through changes in its
trade account resulting from an RMB devaluation, or the real effect first, there are four
channels to analyse:
1. Hong Kong's direct export to China declines, which reduces Hong Kong's
foreign reserve.
2. Hong Kong's re-export (on behalf of the rest of the world) to China may decline.
Assuming that Hong Kong derives some value-added from this entrepot trade (this is
debatable as will be made clear later), it has a negative effect on Hong Kong's
foreign reserve.
3. Hong Kong's direct imports from China may increase, again reducing Hong Kong's
foreign reserve.
S.-J. Wei et al. / China Economic Review 11 (2000) 171±188
177
4. Hong Kong's re-export for China (i.e., China's indirect exports to the rest of the world
through Hong Kong) may increase. Assuming that Hong Kong gets a certain percentage
of value-added from this entrepot trade, its foreign reserve may increase.
The four channels have conflicting effects on Hong Kong's foreign reserve, and hence its
ability to defend the dollar peg (three negative and one positive). This is as far as nonnumerical reasoning can go.
We now turn to simulation results from the CGE model. As a benchmark case, let us
consider a 10% devaluation in RBM (vis-a-vis the US dollar). Table 1A presents the results of
the simulation. Column 2 indicates that Hong Kong's total direct exports to China declines by
US$573 million. Of that, electronics exports fall by US$59 million, textile by US$47 million,
and light manufacturing by US$41 million.
Column 3 of Table 1A shows that Hong Kong's re-exports to China fall by US$3549
million, of which electronics, light manufacturing and textiles fall by US$540 million, $282
million, and $267 million, respectively. As large as these numbers may be in comparison with
Hong Kong's direct exports, the effect on Hong Kong's foreign reserve may be very limited.
Hong Kong's re-export to China, by definition has to be imported from the rest of the world
first. So the effect on Hong Kong's reserve depends on the markup that Hong Kong charges
on this trade.
Feenstra et al. (1998) shows that a reasonable estimate of the markup on this trade is
somewhere between 0% and 0.6%, reflecting an extremely competitive situation in reexporting to China. It has been argued that many mainland companies have branches in Hong
Kong that have as much know how about exporting to China as any Hong Kong firm. This
contributes to an ever lower margin in this business. In any case, if we use the zero markup
estimate, obviously, Hong Kong's reserve is unaffected by the massive reduction in the
indirect export to China (Table 1A, Column 4A). The maximum drop in the foreign reserves
(when the 0.6% estimate is used) is only US$21.3 million (Column 4B), which is less than a
tenth of a percentage of Hong Kong's US$88 billion foreign reserves at the end of the third
quarter in 1998.
After a 10% RMB devaluation, Hong Kong's direct imports from China would increase by
US$925.4 million. More than half of the change come from textile (US$320.1 million), food
and agriculture (US$141.9 million) and ``other manufacturing durables'' (US$228.6 million).
(Table 1A, Column 5).
In contrast, Hong Kong's re-exports for China (i.e., China's indirect export via Hong Kong
to the rest of the world) would increase by US$4046.4 million dollars. Light manufacturing
(US$2564.4 million) and textiles (US$678.5 million) are the two most important items of the
change, reflecting the improved competitiveness of China's labor intensive manufactures after
RMB devaluation. The largest increase is in light manufacturing products, which is composed
mainly of shoes, toys, sports, and travel goods. China's textile products are also competitive,
but its exports are restricted by quotas and competition from other Asia countries. Of course,
just like the re-exports to China by Hong Kong, only the change in the value-added is what
affects Hong Kong's foreign reserve.
According to Feenstra et al. (1998), there are three estimates of the markup for this entrepot
trade: 23%, 27%, and 33%. Feenstra et al.'s (1998) preferred estimate is 27%. In Table 1A,
Percentage of makeupb
Food and agriculture
Mine and energy
Textile
Light manufacturing
Electronics
Motor vehicle and parts
Other manufacturing durables
Total
Import side
Change of
Hong Kong's
direct import
from China
0.23
3
6
156
590
56
2
119
930.67
Change of
Hong Kong's
re-export for
China to rest
of the world
Column 5a
11.65
24.64
678.52
2564.41
241.49
6.74
518.95
4046.40
Markup on
re-export for
China to rest
of the world
Column 6c
ÿ 382.70
ÿ 573.20
141.94
85.49
320.07
100.34
32.59
16.32
228.61
925.36
Column 7Ab
ÿ 365.23
ÿ 196.07
ÿ 266.77
ÿ 281.92
ÿ 540.13
ÿ 192.86
ÿ 1706.40
ÿ 3549.38
0
0
0
0
0
0
0
0
0
0.27
3
7
183
692
65
2
140
1092.53
Markup on
re-export for
China to rest
of the world
Column 7Bb
0.60%
ÿ2
ÿ1
ÿ2
ÿ2
ÿ3
ÿ1
ÿ 10
ÿ 21.30
Export side
ÿ 28.50
ÿ 15.30
ÿ 46.90
ÿ 40.80
ÿ 59.00
Markup on
re-export to
China from
rest of the
world (US$)
Markup on
re-export to
China from
rest of the
world (%)
Change of
Hong Kong's
re-export to
China from rest
of the world
Change of
Hong Kong's
direct export
to China
Percentage of makeupb
Food and agriculture
Mine and energy
Textile
Light manufacturing
Electronics
Motor vehicle and parts
Other manufacturing durables
Total
Column 4Bb
Column 4Ab
Column 3
Column 2a
Sectors
(A) The impact of 10% RMB devaluation on Hong Kong's trade and current account (in million US dollars)
Table 1
0.33
4
8
224
846
80
2
171
1335
Markup on
re-export for
China to rest
of the world
Column 7Cb
178
S.-J. Wei et al. / China Economic Review 11 (2000) 171±188
0% import,
23% export
markup
ÿ 167.76
ÿ 95.12
ÿ 210.91
448.67
ÿ 36.05
ÿ 14.77
ÿ 491.95
ÿ 567.89
Scenarios
Food and agriculture
Mine and energy
Textile
Light manufacturing
Electronics
Motor vehicle and parts
Other manufacturing durables
Total
2ÿ
5+
4A +
7B
ÿ 167.30
ÿ 94.14
ÿ 183.77
551.25
ÿ 26.39
ÿ 14.50
ÿ 471.19
ÿ 406.03
0% import,
27% export
markup
Net change in
forex reserve
due to change
of trade
Column
Column
Column
Column
2ÿ
5+
4A +
7C
ÿ 166.60
ÿ 92.66
ÿ 143.06
705.12
ÿ 11.90
ÿ 14.09
ÿ 440.06
ÿ 163.25
0% import,
33% export
markup
Net change in
forex reserve
due to change
of trade
Column
Column
Column
Column
2ÿ
5+
4B +
7A
ÿ 169.95
ÿ 96.30
ÿ 212.51
446.98
ÿ 39.29
ÿ 15.93
ÿ 502.19
ÿ 589.19
0.6% import,
23% export
markup
Net change in
forex reserve
due to change
of trade
Column
Column
Column
Column
2ÿ
5+
4B +
7B
ÿ 169.49
ÿ 95.31
ÿ 185.37
549.56
ÿ 29.63
ÿ 15.66
ÿ 481.43
ÿ 427.33
0.6% import,
27% export
markup
Net change in
forex reserve
due to change
of trade
Column
Column
Column
Column
2ÿ
5+
4B +
7C
ÿ 168.79
ÿ 93.84
ÿ 144.66
703.42
ÿ 15.14
ÿ 15.25
ÿ 450.30
ÿ 184.55
0.6% import,
33% export
markup
Net change in
forex reserve
due to change
of trade
Column
Column
Column
Column
Using the share of Hong Kong's re-export to China in China's total imports multiply change of China's total imports (exclude direct imports from Hong
Kong) from the model. The share is calculated by the authors based on Hong Kong Custom Statistics and China's total imports at base year (1995) GTAP data.
a
Results directly from the model solution.
b
The markup margin is from Feenstra et al. (1998).
c
Using the share of China's export via Hong Kong in its total exports multiply changes of China's total exports (exclude China's direct export to Hong
Kong) from the model. The share is calculated by authors using export data from the Chinese Customs and China's total exports (exclude Hong Kong's direct
imports from China) at base year (1995) GTAP data.
Net change in
forex reserve
due to change
of trade
2ÿ
5+
4A +
7A
Sectors
Column
Column
Column
Column
(B) The impact on Hong Kong's current account
S.-J. Wei et al. / China Economic Review 11 (2000) 171±188
179
180
S.-J. Wei et al. / China Economic Review 11 (2000) 171±188
Columns 7A±C, we report the changes in the value-added corresponding to each estimate of
the markup.
The net effect of a 10% RMB devaluation depends on the relative size of the above
four effects (Columns 2, plus 4, plus 7 and minus 5). Because there are two markup
estimates for Hong Kong's re-exports to China (from the rest of the world), and three
markup estimates for Hong Kong's re-exports for China (or China's indirect exports),
there are altogether six combinations of the estimates. Table 1B reports the net effect of a
10% RMB devaluation on Hong Kong's foreign reserves through its trade account.
Despite of the increased foreign currency earning for Hong Kong through its entrepot
trading, one sees that all estimates are negative, meaning that the negative effects
dominate the positive effect.
How important is the effect in quantitative term? The six estimates range from ÿ US$163
million to ÿ US$589 million. Hong Kong's trade deficit in 1997 was US$21.4 billion.6 So
the effect of a 10% RMB devaluation is additional increase in Hong Kong's trade deficit (over
the 1997 level) by somewhere between 0.8% to 2.8%, which is relatively small.
To put the estimates into perspective, let us compare them with the actual month-tomonth change in Hong Kong's foreign currency assets. According to a press release of
the Hong Kong Monetary Authority (HKMA), the official foreign currency assets at the
end of April 1998 were US$96.2 billion,7 which was lower by US$600 million from the
end of March figure. The press release states that ``the fall in foreign currency assets held
in the Exchange Fund can be attributed to withdraws of fiscal reserves to meet
government operational needs,'' and that ``monthly figures of the foreign currency assets
are likely to show short-term variations due to seasonal factors.'' So even the largest
estimate of the effect of a 10% RMB devaluation on Hong Kong's official foreign
exchange reserve is smaller than the actual month-to-month fluctuation in the reserve with
an RMB devaluation.
To get a sense of the robustness of the estimates, Tables 2 and 3 report simulation results
for a 5% and 20% devaluation of the Chinese RMB. We perform exactly the same set of
calculations as in Table 1. In both instances, Hong Kong would see a large increase in its
foreign exchange earning through value-added derived from its re-exports for China to the
rest of the world. But this increase in the foreign exchange earning is not big enough to
offset the decrease in earning through its widened trade deficit in its direct trade with China.
After a 5% RMB devaluation, the net decline in Hong Kong's foreign currency assets will
range from US$119 million to US$328 million. After a 20% RMB devaluation, the net
decline in Hong Kong's foreign currency assets will be between US$327 million and
US$1158 million. These are relatively small effects compared with the actual size of Hong
6
According to the table labeled as ``Gross Domestic Product Estimates by Expenditure Component,'' in
HKMA's home page (www.info.gov.hk/hkma), Hong Kong's total exports of goods (fob) and total imports of
goods (cif) in 1997 were Hong Kong $1403.9 billion and $1532.9 billion (current price), respectively. Using an
exchange rate of US$1 = Hong Kong$7.74, this is equivalent to a trade deficit of US$21.4 billion.
7
``Hong Kong's Latest Foreign Currency Assets Figures Released,'' in HKMA's home page (www.info.
gov.hk/hkma/new/press/exchangefund/980522e.htm).
S.-J. Wei et al. / China Economic Review 11 (2000) 171±188
181
Table 2
(A) The impact of 5% RMB devaluation on Hong Kong's trade and current account (in million US dollars)
Sectors
Export side
Percentage of
makeupb
Food and
agriculture
Mine and energy
Textile
Light
manufacturing
Electronics
Motor vehicle
and parts
Other
manufacturing
durables
Total
Import side
Percentage of
makeupb
Food and
agriculture
Mine and energy
Textile
Light
manufacturing
Electronics
Motor vehicle
and parts
Other
manufacturing
durables
Total
Column 2a
Column 3
Column 4Ab
Column 4Bb
Change of
Hong Kong's
direct export
to China
Change of
Hong Kong's
re-export to
China from
rest of the
world
Markup on
re-export to
China from
rest of the
world (%)
Markup on
re-export to
China from
rest of the
world (US$)
0
0.60%
ÿ 14.87
ÿ 215.14
0
ÿ1
ÿ 8.13
ÿ 25.56
ÿ 21.49
ÿ 113.83
ÿ 152.36
ÿ 156.92
0
0
0
ÿ1
ÿ1
ÿ1
ÿ 30.92
ÿ 0.02
ÿ 296.34
ÿ 102.36
0
0
ÿ2
ÿ1
ÿ 199.68
ÿ 946.51
0
ÿ6
ÿ 300.67
ÿ 1983.46
0
ÿ 11.90
Column 5a
Column 6c
Column 7Ab
Column 7Bb
Column 7Cb
Change of
Hong Kong's
direct import
from China
Change of
Hong Kong's
re-export for
China to rest
of the world
Markup on
re-export for
China to rest
of the world
Markup on
re-export for
China to rest
of the world
Markup on
I re-export
for China to
rest of the
world
0.23
0.27
0.33
71.81
5.25
1
1
2
43.14
164.26
50.51
11.26
313.80
1269.96
3
72
292
3
85
343
4
104
419
16.52
7.87
120.09
2.81
28
1
32
1
40
1
114.94
249.67
57
67
82
469.05
1972.85
453.76
532.67
651
(continued on next page)
182
S.-J. Wei et al. / China Economic Review 11 (2000) 171±188
Table 2 (continued )
(B) The impact on Hong Kong's current account
Column
Column
Column
Column
2ÿ
5+
4A +
7A
Column
Column
Column
Column
2ÿ
5 +
4A +
7B
Column
Column
Column
Column
2ÿ
5+
4A +
7C
Column
Column
Column
Column
2ÿ
5+
4B +
7A
Column
Column
Column
Column
2ÿ
5+
4B +
7B
Column
Column
Column
Column
2ÿ
5+
4B +
7C
Sectors
Net change in
forex reserve
due to change
of trade
Net change in
forex reserve
due to change
of trade
Net change in
forex reserve
due to change
of trade
Net change in
forex reserve
due to change
of trade
Net change in
forex reserve
due to change
of trade
Net change in
forex reserve
due to change
of trade
Scenarios
0% import,
23% export
markup
0% import,
27% export
markup
0% import,
33% export
markup
0.6% import,
23% export
markup
0.6% import,
27% export
markup
0.6% import,
33% export
markup
ÿ 85.26
ÿ 84.95
ÿ 86.76
ÿ 86.55
ÿ 86.24
ÿ 48.23
ÿ 105.10
270.89
ÿ 47.55
ÿ 86.27
347.09
ÿ 49.36
ÿ 118.56
219.15
ÿ 48.91
ÿ 106.01
269.95
ÿ 48.24
ÿ 87.18
346.15
ÿ 15.02
ÿ 7.13
ÿ 7.81
ÿ 6.96
ÿ 21.60
ÿ 7.86
ÿ 16.79
ÿ 7.74
ÿ 9.59
ÿ 7.58
ÿ 247.21
ÿ 232.23
ÿ 262.87
ÿ 252.89
ÿ 237.91
0.00
0.00
0.00
0.00
0.00
0.00
ÿ 237.05
0.00
ÿ 118.68
0.00
ÿ 327.87
0.00
ÿ 248.95
0.00
ÿ 130.58
Food and
ÿ 85.47
agriculture
Mine and energy ÿ 48.68
Textile
ÿ 117.65
Light
220.09
manufacturing
Electronics
ÿ 19.82
Motor vehicle
ÿ 7.24
and parts
ÿ 257.19
Other
manufacturing
durables
Housing and
0.00
construction
Services
0.00
Total
ÿ 315.96
Using the share of Hong Kong's re-export to China in China's total imports multiply change of China's total
imports (exclude direct imports from Hong Kong) from the model. The share is calculated by the authors based on
Hong Kong Custom Statistics and China's total imports at base year (1995) GTAP data.
a
Results directly from the model solution.
b
The markup margin is from Feenstra et al. (1998).
c
Using the share of China's export via Hong Kong in its total exports multiply changes of China's total
exports (exclude China's direct export to Hong Kong) from the model. The share is calculated by authors using
export data from the Chinese Customs and China's total exports (exclude Hong Kong's direct imports from China)
at base year (1995) GTAP data.
Kong's trade deficit, or the actual size of its official foreign assets, or the month-to-month
movement in the official foreign assets.
4. The psychological factor
Section 3 shows that a devaluation of the Chinese currency will have a small negative
effect on Hong Kong's foreign exchange reserve through its effect on its trade account. As we
stated earlier, the speculative pressure on Hong Kong dollar comes from the market
S.-J. Wei et al. / China Economic Review 11 (2000) 171±188
183
Table 3
(A) The impact of 20% RMB devaluation on Hong Kong's trade and current account (in million US dollars)
Sectors
Column 2a
Column 3
Column 4Ab
Column 4Bb
Export side
Change of
Hong Kong's
direct export
to China
Change of
Hong Kong's
re-export to
China from
rest of the world
Markup on
re-export to
China from
rest of the
world (%)
Markup on
re-export to
China from
rest of the
world (US$)
Percentage of
makeupb
Food and
agriculture
Mine and energy
Textile
Light
manufacturing
Electronics
Motor vehicle
and parts
Other
manufacturing
durables
Total
Import side
Percentage of
makeupb
Food and
agriculture
Mine and energy
Textile
Light
manufacturing
Electronics
Motor vehicle
and parts
Other
manufacturing
durables
Total
0
0.60%
ÿ 58.20
ÿ 840.82
0
ÿ5
ÿ 31.68
ÿ 98.03
ÿ 83.75
ÿ 443.79
ÿ 585.01
ÿ 611.12
0
0
0
ÿ3
ÿ4
ÿ4
ÿ 120.84
ÿ 0.08
ÿ 1103.55
ÿ 400.97
0
0
ÿ7
ÿ2
ÿ 660.14
ÿ 3704.59
0
ÿ 22
ÿ 1052.72
ÿ 7689.85
0
ÿ 46.14
Column 5a
Column 6c
Column 7Ab
Column 7Bb
Column 7Cb
Change of
Hong Kong's
direct import
from China
Change of
Hong Kong's
re-export for
China to rest
of the world
Markup on
re-export for
China to rest
of the world
Markup on
re-export for
China to rest
of the world
Markup on
I re-export
for China to
rest of the
world
0.23
0.27
0.33
285.56
21.07
5
6
7
171.77
648.59
201.36
44.96
1246.30
5049.39
10
287
1161
12
337
1363
15
411
1666
65.63
32.06
476.57
11.45
110
3
129
3
157
4
458.49
996.18
229
269
329
1863.46
7845.92
1804.56
2118.40
2589
(continued on next page)
184
S.-J. Wei et al. / China Economic Review 11 (2000) 171±188
Table 3 (continued)
(B) The impact on Hong Kong's current account
Column
Column
Column
Column
2ÿ
5+
4A +
7A
Column
Column
Column
Column
2ÿ
5+
4A +
7B
Column
Column
Column
Column
2ÿ
5+
4A +
7C
Column
Column
Column
Column
2ÿ
5+
4B +
7A
Column
Column
Column
Column
2ÿ
5+
4B +
7B
Column
Column
Column
Column
2ÿ
5+
4B +
7C
Sectors
Net change in
forex reserve
due to change
of trade
Net change in
forex reserve
due to change
of trade
Net change in
forex reserve
due to change
of trade
Net change in
forex reserve
due to change
of trade
Net change in
forex reserve
due to change
of trade
Net change in
forex reserve
due to change
of trade
Scenarios
0% import,
23% export
markup
0% import,
27% export
markup
0% import,
33% export
markup
0.6% import,
23% export
markup
0.6% import,
27% export
markup
0.6% import,
33% export
markup
ÿ 338.07
ÿ 336.81
ÿ 343.96
ÿ 343.12
ÿ 341.85
ÿ 191.31
ÿ 410.12
1078.22
ÿ 188.61
ÿ 335.34
1381.19
ÿ 195.77
ÿ 463.48
872.58
ÿ 193.97
ÿ 413.63
1074.56
ÿ 191.28
ÿ 338.85
1377.52
ÿ 57.79
ÿ 29.05
ÿ 29.20
ÿ 28.36
ÿ 83.48
ÿ 31.91
ÿ 64.42
ÿ 31.45
ÿ 35.82
ÿ 30.77
ÿ 849.66
ÿ 789.89
ÿ 911.74
ÿ 871.89
ÿ 812.12
ÿ 797.78
ÿ 327.03
ÿ 1157.76
ÿ 843.92
ÿ 373.17
ÿ 338.91
Food and
agriculture
Mine and energy ÿ 193.11
Textile
ÿ 459.97
876.25
Light
manufacturing
Electronics
ÿ 76.86
Motor vehicle
ÿ 29.51
and parts
ÿ 889.51
Other
manufacturing
durables
Total
ÿ 1111.62
Using the share of Hong Kong's re-export to China in China's total imports multiply change of China's total
imports (exclude direct imports from Hong Kong) from the model. The share is calculated by the authors based on
Hong Kong Custom Statistics and China's total imports at base year (1995) GTAP data.
a
Results directly from the model solution.
b
The markup margin is from Feenstra et al. (1998).
c
Using the share of China's export via Hong Kong in its total exports multiply changes of China's total
exports (exclude China's direct export to Hong Kong) from the model. The share is calculated by authors using
export data from the Chinese Customs and China's total exports (exclude Hong Kong's direct imports from China)
at base year (1995) GTAP data.
perception of the Hong Kong authority's commitment to defend the currency as much as from
the perception of the authority's ability to defend the currency. How would the market
participants in Hong Kong react to a (hypothetical) news of RMB devaluation?
Towards this end, a survey of seven market participants was conducted.8 The respondents
include a currency trader, two securities traders, two portfolio (mutual fund) managers, and
two investment bank economists. The survey was conducted between March 15 and March
25, 1998 in Hong Kong. The survey consists of three substantive questions in addition to the
8
We would like to thank Simone S. of an investment bank in Hong Kong for help in carrying out the survey.
S.-J. Wei et al. / China Economic Review 11 (2000) 171±188
185
question asking the respondent to identify the line of business he or she is in. Because the
survey is relatively short, we repeat the questions below.
Question 1: How much do you think the Chinese RMB is overvalued?
a. Over 20%
b. Between 10% and 20%
c. Below 10%
d. The current rate is about right.
e. It is somewhat undervalued.
f. Do not know
Question 2: One hypothesis says that if the RMB devalues (depreciates), Hong Kong's
exports will become more competitive in the world market because most of the goods that
Hong Kong exports are made on the mainland. As a result, Hong Kong's foreign exchange
reserve will again start to rise. Hong Kong will be in a better position to defend its US dollarlink exchange rate system. Do you agree with this hypothesis?
a. Strongly agree
b. Basically agree
c. Basically disagree
d. Strongly disagree
e. Do not know
Question 3: Another hypothesis says that if the RMB devalues, currency traders and
ordinary people in Hong Kong will immediately start to sell Hong Kong dollars possibly in
panic. They do so because they will lose confidence in the ability of HKMA to defend the
Hong Kong dollar despite possible benefits that the RMB devaluation may have on Hong
Kong. Do you agree with this hypothesis?
a. Strongly agree
b. Basically agree
c. Basically disagree
d. Strongly disagree
e. Do not know
Table 4 presents the results of the survey. On the first question, two respondents thought
that the RMB (vs. US dollar) exchange rate was about right, two thought it was overvalued
between 10% and 20%, and three were not sure either way. In other words, there was no
strong sentiment that the RMB is severely overvalued. This ambiguity in the answers is
perhaps not surprising. From 1994 to mid-1997, the Chinese inflation rate had been a lot
higher than the US rate. Yet, its currency, under a (tightly) managed float system, was actually
appreciating in value at about 1% year. Since mid-1997, its inflation rate has come down
dramatically, to almost zero (in terms of CPI), or negative inflation (in terms of retail prices),
in the first quarter of 1998.
186
S.-J. Wei et al. / China Economic Review 11 (2000) 171±188
Table 4
Survey of market participants
Currency
trader
Security
trader
Question 1: How much is RMB overvalued?
> 20%
Between 10% and 20%
y
< 10%
About right
y
Somewhat undervalued
Do not know
Security
trader
Fund
manager
Fund
manager
Economist
Economist
y
y
y
y
Question 2: An RMB devaluation will increase Hong Kong's foreign exchange reserve
Strongly agree
Basically agree
y
y
Basically disagree
y
y
y
Strongly disagree
y
Do not know
Question 3: An RMB devaluation will lead to panic selling of the Hong Kong dollar
Strongly agree
y
y
Basically agree
y
y
y
y
Basically disagree
Strongly disagree
Do not know
y
y
y
Notes: (1) The survey was conducted in Hong Kong between March 15 and March 25, 1998. The full
questions can be found in the text of the paper. (2) ``y'' indicates a check.
On the second question, five out of seven people disagree with the hypothesis that an RMB
devaluation will raise Hong Kong's foreign exchange reserve. But most did not check
``strongly disagree.'' Two respondents actually checked ``basically agree.'' This pattern of
answers corresponds well to our discussion in Section 3 that the effect of an RMB
devaluation has four effects on Hong Kong's trade account, in conflicting directions. While
the net effect from the CGE simulations is negative, it is quantitatively small, which might
explain why the survey respondents (without the tool of a CGE model) do not have a
consensus on this.
The responses to the third question stand out in sharp contrast to those of the early two
questions. All respondents agree with the hypothesis that an RMB devaluation will lead to a
panic selling of the Hong Kong dollar. This suggests that the effect of an RMB devaluation on
the Hong Kong dollar, in people's mind, must go beyond what it does to Hong Kong's trade
account (or even foreign exchange reserves).
We should be careful about drawing too much from the survey given its small sample
size. The small sample prevents us from doing formal statistical testing. On the other hand,
even in this small sample, it is remarkable that all respondents agree on the psychological
effect of an RMB devaluation on the Hong Kong dollar (i.e., possible panic selling will
be triggered).
S.-J. Wei et al. / China Economic Review 11 (2000) 171±188
187
5. Summary and concluding remarks
A devaluation of the Chinese yuan will affect the defensibility of the peg of the Hong
Kong dollar vis-a-vis the US dollar through its impact on Hong Kong's trade balance, and its
impact on the psychology of foreign exchange market participants. Hong Kong's trade
balance is shaped to a significant extent by Hong Kong being the conduit of a huge amount of
indirect trade between China and the rest of the world. It derives a significant income from
the markup on its indirect exports for China but virtually no income from its indirect exports
to China. This observation seems to suggest that a devaluation of the Chinese currency could
increase Hong Kong's foreign exchange earnings. Indeed, our CGE simulation shows that
this effect is present. However, Hong Kong also loses foreign exchange earnings through its
direct trade with China. The latter, negative effect, dominates the former, positive effect. On
the other hand, the net effect is quantitatively small compared to Hong Kong's actual trade
deficit or the actual size of its official foreign assets or the month-to-month fluctuations in the
official foreign assets.
The psychological impact of a yuan devaluation is studied through a (small-sample)
survey of financial market participants in Hong Kong in March 1998. The survey revealed
several pieces of interesting information. The respondents were not sure whether the
Chinese yuan was overvalued relative to the US dollar. While most of the respondents
thought that an RMB devaluation would not increase Hong Kong's foreign exchange
reserves, everyone believed that an RMB devaluation would lead to a panic selling of the
Hong Kong dollar.
The limited evidence that we have presented suggests that a yuan devaluation would tend
to trigger a speculative attack on the Hong Kong dollar. Although our CGE model indicates
that a yuan devaluation is likely to have a negligible effect on Hong Kong's overall trade
balance because of four offsetting channels, the market psychology may work against the
Hong Kong dollar.
Acknowledgments
The research for the paper was supported in part by a grant from the Japan's Ministry of
Finance to Harvard Institute for International Development. The views in the paper are those
of the authors, and may not be shared by any of the institutions that they are or have been
associated with.
References
Feenstra, R. C., Hai, W., Woo, W. T., Sachs, J. D., & Yao, S. (1998). The U.S. ± China bilateral trade balance: its
size and determinants. Paper presented at the UNDP ± HIID Conference on China's Integration into the Global
Economy. Harvard University, January 17.
Howe, H. (1975). Development of the extended linear expenditure system from simple savings assumptions. Eur
Econ Rev 6(3), 305 ± 310 (July).
Hu, F. (1998). China's currency is safe. Far East Econ Rev February 5.
188
S.-J. Wei et al. / China Economic Review 11 (2000) 171±188
Hu, F., Li, D., & Li, S. (1998). Yuan devaluation: a paper tiger. Asian Wall St J February 19.
Noland, M., Liu, L., Robinson, S., & Wang, Z. (1998). The Asian financial crisis and global adjustment. International Economics Policy Analysis Series 56. Washington, DC: Institute for International Economics.
Wang, Z. (1997a). China and Taiwan access to the World Trade Organization: implications for U.S. agriculture
and trade. Agric Econ 17, 239 ± 264 (December).
Wang, Z. (1997b). The impact of China and Taiwan joining the World Trade Organization on U.S. and world
agricultural trade: a computable general equilibrium analysis. Technical Bulletin. No. 1858, US Department of
Agriculture, Economic Research Service.
Wang, Z., & Kinsey, J. (1994). Consumption and saving behavior under strict and partial rationing. China Econ
Rev Ð Int J 5(1), 83 ± 100 (Spring).
Wei, S.-J. (1998). Why does China receive so little foreign direct investment? Working Paper. Kennedy School,
Harvard University.
Wei, S.-J., & Zeckhauser, R. (1998). Two crisis and two Chinas. Jpn World Econ 10(3), 359 ± 369 (July).